Firefly - Audio/VideoInstall

Summary

Transcribe audio or video media from a pre-signed URL to text, optionally translating to specified target languages and generating caption files.

Inputs

• Input Type (required) — Select whether the input media is audio or video.
• Source URL (required) — Pre-signed URL pointing to the audio or video file to transcribe.
• Audio Media Type — Specify the audio file's media type if input type is audio.
• Video Media Type — Specify the video file's media type if input type is video.
• Target Locale Codes — Comma-separated list of language codes for translation; leave empty to transcribe in the source language.
• Generate Captions — Set to true to generate SRT caption files alongside the transcription.

Output shape

a single JSON object per input item containing the transcription results and optional captions

Output includes transcribed text in the source or target languages depending on input. Caption files are included if enabled. There is no pagination; each input item results in one output item. Errors are returned per item if continueOnFail is set.

Examples

Example 1: Transcribe a video file in MP4 format located at a pre-signed URL, producing English transcript and optionally captions.

Input Type: video; Source URL: https://presigned-url.com/media.mp4; Video Media Type: video/mp4; Target Locale Codes: (empty for no translation); Generate Captions: true/false.